Introduction
Alzheimer’s disease (AD) is a severe multifactorial disorder leading to progressive dementia and eventually death. One of the major hallmarks of AD is the accumulation of soluble beta-amyloid (Aβ) peptides within the brain and the deposition of fibrillar forms of Aβ extracellularly [
1]. The accumulation of fibrillar Aβ-containing plaques leads to the proinflammatory activation of microglia and astrocytes that surround the deposits. The ineffective clearance of the fibrillar Aβ deposits results in their chronic production of cytotoxic factors that act to exacerbate AD-like pathology and neuronal death [
2,
3].
Peroxisome proliferator-activated receptors (PPARs) are ligand-activated transcription factors that regulate cellular metabolism by binding to sequence-specific DNA elements. There are 3 PPAR isoforms, α, β/δ (hereafter referred to as PPARδ) and γ. In general, PPARs are lipid sensors and principally regulate fatty acid and cholesterol metabolism [
4,
5]. Importantly, PPARs act to regulate inflammatory processes in microglia and macrophages, suppressing the elaboration of cytokines and other inflammatory mediators and promoting tissue repair and phagocytosis [
6,
7]. PPARγ has been intensively studied in mouse models of central nervous system (CNS) diseases, including AD, in which its activation has been shown to lead to improvement in learning and memory and concomitant amelioration of AD-like pathology [
8-
11]. However, PPARδ is far less studied in models of brain diseases. PPARδ agonists have been shown to reduce the production of inflammatory mediators [
12] especially in peripheral immune cells. PPARδ ligands have been shown to be neuroprotective in
in vivo models of Parkinson’s disease (PD) [
13,
14], brain ischemia [
14,
15], spinal cord injury [
16] and in streptozotocin-induced experimental diabetes [
17]. Thus far, only a single study has addressed the effects of PPARδ activation in a mouse model of AD, showing that 1-month treatment of 5XFAD mice with PPARδ agonist GW0742 led to reduction in brain Aβ burden, reduced astrocytic activation and increased expression of Aβ-degrading enzymes [
18]. Since the 5XFAD mice exhibit age-related neuronal degeneration in specific brain areas, we wished to dissect the protective effect PPARδ activation in 5XFAD mice in more detail, focusing especially on inflammation and AD-related neuronal death. Here we show that a 2-week oral treatment of 5XFAD mice with GW0742 reduced the brain Aβ load and microglial activation without affecting the number of neurons containing intracellular Aβ/amyloid precursor protein (APP). Importantly, we show for the first time that the treatment attenuated the degeneration of neurons in the subiculum of the 5XFAD mice. GW0742 was effective in preventing lipopolysaccharide (LPS)-induced increase in inflammatory mediators in primary microglia
in vitro. Whereas GW0742 alone failed to prevent neuron loss against glutamate exposure, it significantly increased neuronal survival in inflammation-induced neuron death
in vitro. Our data demonstrate that GW0742 is a powerful anti-inflammatory agent with neuroprotective properties and PPARδ agonism could be considered as a potential AD therapy.
Materials and methods
Animals and drug treatment
5XFAD male mice, originally described by Oakley et al., were a gift from Dr. Robert Vassar (Northwestern University) and B6SJL/F1 females were purchased from Jackson Laboratories (Bar Harbor, ME, USA). A total of 32 mice, both males and females were used in this study. The mice were randomized into study groups: wild type (WT) vehicle- treated: 9 female and 5 male mice; transgenic (TG) vehicle-treated: 4 female and 5 male mice and TG GW0742-treated: 5 female and 4 male mice. GW0742 was provided by GlaxoSmithKline (Research Triangle, NC, USA). GW0742 was given by oral gavage at the dose of 30 mg/kg daily as a water suspension for 2 weeks starting at the age of 4.5 months. Vehicle-treated mice received water only. The animals were sacrificed at the end of the treatment period 6 hours after the last dose of GW0742. All animal experimentation was done according to the Case Western Reserve University Institutional Animal Care and Use Committee guidelines.
Immunohistochemistry
At the end of the treatment period the mice were anesthetized with Avertin and transcardially perfused with 0.01 M, PBS pH 7.4. Brains were removed, and the right hemisphere was immersion fixed with 4% PFA in 0.1 M phosphate buffer (PB, pH 7.4) over night at 4°C. Thereafter, the brains were cryoprotected with 10% sucrose for 24 hours following incubation in 30% sucrose for 48 hours after which the brains were frozen and cut in serial 10-μM sagittal sections. The sections were incubated with antibodies to glial fibrillary acidic protein (GFAP; 1:500 dilution, Dako, Carpinteria, CA, USA), ionized calcium binding adaptor molecule 1 (Iba-1; 1:200 dilution, Wako Chemicals, Richmond, VA, USA), NeuN (Aves Labs Inc, Tigard, Oregon, USA), complement component 3 (C3) and C1qa (1:1,000 dilution, both from Novus Biologicals, Littleton, CO, USA,), and 6E10 (BioLegend, Dedham, MA, USA) followed by incubation with appropriate Alexafluor 488 or 546 conjugated secondary antibodies (Molecular Probes/Life Technologies, Grand Island, NY, USA).
Images of the hippocampi were taken from 3 sections per animal, approximately 1,200 μM apart and spanning the hippocampi. Immunoreactivity was quantified by using ImagePro Premium (Media Cybernetics, Rockville, MD, USA) blinded to the study groups and presented as percentage of positively-stained area in the hippocampi. NeuN positive neurons were counted in the subiculum region of the hippocampi in two sections from each animal. The cell count data weres confirmed by quantifying the percentage of NeuN immunoreactive area in the subiculum area of the hippocampi. Cortical 6E10 immunoreactive neuronal bodies were counted and cortical 6E10 immunoreactivity was quantified from a total of 3 separate images taken from layer V in the cortex. Intensity of the intraneuronal 6E10 immunoreactivity was quantified using ImageJ by outlining 10 to 14 individual 6E10 immunopositive neurons from 3 separate images of cortical layer V taken from 3 consecutive sections.
Primary cortical neuronal cultures
Primary neurons were cultivated as described previously [
19]. Briefly, cortices of embryonic day 15 C57BL/6J pups were dissected and freed from their meninges. After dissociation with 0.025% (w/v) trypsin in Krebs buffer (0.126 M NaCl, 2.5 mM KCl, 25 mM NaHCO
3, 1.2 mM NaH
2PO
4, 1.2 mM MgCl
2, 2.5 mM CaCl
2, pH 7.4) for 20 minutes at 37°C the tissues were treated with 0.008% w/v DNaseI and 0.026% w/v trypsin inhibitor (Sigma, St. Louis, MO, USA) and centrifuged at 256 × g for 3 minutes. The cell pellet was resuspended in 3 ml of DNaseI/SBT1 (Sigma, St. Louis, MO, USA) in Krebs solution and gently triturated through a blunt-ended glass pipet. Seven milliliters of additional Krebs buffer were added, the cell suspension centrifuged at 256 × g for 3 minutes and the cells were resuspended in Neurobasal Medium (Gibco/Life Technologies, Grand Island, NY, USA) supplemented with 0.2 mM L-glutamine (Gibco, Grand Island, NY, USA), 0.01 mg/ml gentamicin (Sigma, St. Louis, MO, USA) and B27 Supplement (Gibco, Grand Island, NY, USA), filtered through a 200 μM nylon mesh filter and counted using a hemocytometer. Primary cortical neurons were plated onto poly-D-lysine (50 μg/ml in water) and laminin (5 μg/ml water; Sigma, St. Louis, MO, USA) 24-well plates at the density of 200,000 cells per well. After 5 to 6 days
in vitro the cells were pre-exposed to 1 μM GW0742 for 6 hours followed by exposure to 500 μM glutamate in the presence of 1 μM GW0742 for 24 hours. Cell viability was measured by MTT assay. To assess the effect of 1 μM GW0742 in neuronal viability, the cells were exposed to 1 μM GW0742 alone.
Primary microglia cultures
Primary microglia were cultivated as described previously by using mild trypsinization [
11]. Briefly, P0-P3 C57BL/6J mouse pups were decapitated, the brains removed, rinsed with PBS containing 1 g/l glucose, mechanically dissociated and digested with 0.5% trypsin-EDTA for 20 minutes at 37°C. Thereafter, the tissue homogenate was resuspended in DMEM/F12 media (Gibco, Grand Island, NY, USA) containing 10% heat-inactivated FBS (Gibco, Grand Island, NY, USA) and 1% penicillin-streptomycin. After trituration the cell suspension was plated onto 150 mm culture dishes for 20 to 22 days at 37°C and 5% CO
2. After the plates were confluent, astrocytes were removed by incubating the plates with 0.25% trypsin in Hank's Balanced Salt Solution (HBSS) diluted in 1:4 serum-free DMEM/F12 for 30 minutes to 1 hour at 37°C. After washing the plates with PBS, microglia attached on the plates were removed by trypsinization with 0.25% trypsin in PBS. The action of trypsin was stopped with DMEM/F12 media/10% heat-inactivated FBS, cells centrifuged and plated for subsequent studies. To analyze the effect of GW0742 on primary microglial viability, microglia cultures were exposed to 1 μM GW0742 for 24 hours and cell survival was analyzed by MTT assay.
Primary neuron-microglia co-cultures
Primary neuron-microglia co-cultures were prepared as described by Gresa-Arribas
et al. [
20]. Briefly, primary cortical neurons were plated onto 24-well plate. At 5 to 6 days
in vitro primary microglia were isolated and plated on the top of neurons in Neurobasal Medium (Gibco, Grand Island, NY, USA) supplemented with 0.2 mM L-glutamine (Gibco, Grand Island, NY, USA), 0.01 mg/ml gentamicin (Sigma, St. Louis, MO, USA) and B27 Supplement (Gibco, Grand Island, NY, USA) at the density of 1:2 (100,000 microglia per 200,000 neurons). The next day the co-cultures were exposed to 1 μM GW0742 for 6 hours after which they were exposed to 100 ng/ml LPS and 30 ng/ml interferon (IFN)γ (Preprotech, Rocky Hill, NJ, USA) for 48 hours. The cells were rinsed with PBS (pH 7.4), fixed with 4% PFA for 20 minutes, permeabilized with 0.2% Triton-x in PBS for 10 minutes and incubated with an anti-microtubule-associated protein 2 (MAP-2) antibody (Sigma, St. Louis, MO, USA) in 5% NGS following incubation with Alexa-488 conjugated secondary antibody (Molecular Probes, Eugene, OR, USA). Neuronal viability was evaluated by quantifying the extent of MAP-2 immunoreactivity in the microglia-neuron co-cultures. MAP-2 immunoreactivity reveals any alterations both in the dendritic compartment and the cell soma and is frequently used to assess neuronal integrity and viability in co-culture systems [
20].
Tissue dissection
At the time of sacrifice, the animals were terminally anesthetized with Avertin and perfused with PBS. Brains were removed and cortices dissected out. Hemibrains were homogenized in 800 μl of tissue homogenization buffer (250 mM sucrose, 20 mM Tris, 1 mM ethylenediaminetetraacetic acid (EDTA), 1 mM ethylene glycol tetraacetic acid (EGTA) in diethylpyrocarbonate-treated water) containing Protease Inhibitor Cocktail (1:100, Sigma, St. Louis, MO, USA). The homogenates were centrifuged at 5,000 × g for 10 minutes at 4°C and supernatants stored at −80°C and used for Western blot analysis.
RT-PCR
For RT-PCR, primary microglia were plated at the density of 1 × 106 cells per well and stimulated in serum-free DMEM/12 (Gibco, Grand Island, NY, USA) for 24 hours prior to stimulation with 1 μM GW0742 for 24 hours followed by 10 ng/ml LPS (Sigma, St. Louis, MO, USA) together with the GW0742 for 24 hours. Thereafter, the plates were washed with PBS and mRNA isolated using RNeasy Mini kit (Qiagen, Valencia, CA, USA) according to the manufacturer’s instructions.
Cortical brain samples were homogenized in homogenization buffer and an equivalent amount of RNABee (TelTest Inc, Friendwood, TX, USA) was added to the samples. Thereafter, 0.2 ml of chloroform (Sigma, St. Louis, MO, USA) were added, the samples centrifuged for 15 minutes at 13,000 × g at 4°C, equal amount of 70% ethanol was added to the aqueous layer followed by mRNA isolation using RNease Mini kit (Qiagen, Valencia, CA, USA). mRNA concentration and purity was determined using a NanoDrop 2000 (Thermo Scientific, Hudson, NH, USA). Equivalent amounts of mRNA were reverse transcribed using a QuantiTect Reverse Transcription kit (Qiagen, Valencia, CA, USA) according to the manufacturer’s instructions. The procedure included elimination of genomic DNA. The cDNA was preamplified for 14 cycles using a TagMan PreAmp Master Mix for select primer sets (Applied Biosystems/Life Technologies, Foster City, CA, USA). Quantitive PCR was performed with the StepOne Plus Real Time PCR system (Applied Biosystems, Foster City, CA, USA) for 40 cycles. Analysis of gene expression was performed using the comparative C
t method (ΔΔC
T) where the threshold cycle for the target genes was normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and rRNA internal housekeeping gene controls (ΔC
T). The mRNA expression was presented as fold change and statistical analyses were performed on ΔC
T ± SEM for each target gene as described earlier [
21].
Western blotting
Protein concentration of the brain lysates was determined by BCA (Pierce, Rockford, IL, USA). Equal amounts of protein were run on Bis-Tris 4 to 12% gels (Life Technologies, Foster City, CA, USA). The following antibodies were used: anti-actin (Santa Cruz Biotechnology, Dallas, TX, USA); anti-apolipoprotein E (ApoE) (Santa Cruz Biotechnology, Dallas, TX, USA); anti-β-actin (Santa Cruz Biotechnology, Dallas, TX, USA); anti-ATP-binding cassette transporter A1 (Abca1; Novus Biologicals, Littleton, CO, USA) and G1 (Abcg1; Novus Biologicals, Littleton, CO, USA) followed by incubation with horseradish peroxidase (HRP)-conjugated secondary antibodies.
Discussion
Here we report that PPARδ activation in an animal model of AD results in reduction in the extracellular plaque burden that is associated with a robust reduction in inflammation. Importantly, treatment with a PPARδ agonist provided robust neuroprotection, with a significant attenuation of the loss of neurons in the subiculum of 5XFAD mice. Neuronal death is one of the main pathological features in patients with AD. The current study is thus far the only study linking PPARδ-mediated reduction in the Aβ load with preservation of neurons in a transgenic mouse model of AD. Importantly, our study shows that the protection of the neurons was not associated with a reduction in intraneuronal Aβ by PPARδ activation, but rather with a suppression of inflammation. Of importance is that a significant reduction in the levels of plaque-associated 6E10 immunoreactivity was achieved with a brief 2-week treatment period. The fact that the amount of Aβ-containing neurons and their 6E10 immunoreactivity was not altered upon GW0742 treatment suggest that our GW0742 treatment paradigm had a direct effect on microglia and did not affect the accumulation of intraneuronal Aβ.
An earlier study showed that PPARδ activation led to decreased levels of 6E10 immunoreactivity only in the subiculum of 5XFAD mice [
18]. This is in accordance with our data showing decreased levels of 6E10 immunoreactivity in the subiculum, but in addition we also noted a significant decrease in the levels of Aβ deposits in hippocampus and in cortical layer V. The reduction was specific to extracellular deposits since the number of 6E10 positive neurons remained unchanged upon treatment. There are three major differences between the current study and the study by Kalinin
et al. First, the age of the animals at the start of the treatment in our study was 4.5 months compared to 2 or 3.5-month-old mice used in the study by Kalinin
et al. The second is the treatment time, which was only 2-weeks in the current study compared to 1-month period in the study by Kalinin
et al. Lastly, we administered the drug by oral gavage once a day, whereas Kalinin
et al. provided the drug in chow. Our study shows that PPARδ agonist GW0742 is very effective in reducing the levels of extracellular Aβ deposits over a relatively short treatment period. In contrast to Kalinin
et al., the mechanism underlying the clearance of Aβ was not mediated through increased neprilysin (NEP) or insulin-degrading enzyme (IDE) expression, as we were unable to detect GW0742-stimulated mRNA expression levels of NEP and IDE (data not shown). Instead, we postulate that the reduction is due to enhanced microglial-mediated clearance of Aβ.
A large body of evidence suggests that ApoE is one of the main mechanisms underlying soluble Aβ clearance within the interstitial fluid in the AD brain [
9,
10] and is involved in microglia-mediated degradation of soluble Aβ [
8-
11,
27]. Since some studies have suggested that PPARδ agonists induce the expression of Abca1 [
28,
29], we wanted to analyze whether the reduction in the Aβ deposition was due to increased expression of Abca1 or Abcg1 and subsequent increase in ApoE lipidation. Our results show that GW0742 treatment unexpectedly reduced the protein expression levels of these proteins. Whilst this may be attributed to differences in the models used in previous studies [
28,
29], it is important to note that the levels of total ApoE and lipidated ApoE were not altered by GW0742 suggesting that the effect of GW0742 is not mediated through induction of ApoE lipidation. This is consistent with published literature as no data showing elevation in the levels of ApoE upon PPARδ activation have been reported to date in this context, nor are predicted from the sequence of the nuclear receptor response element in the ApoE promoter.
In the current study GW0742 treatment reduced the expression levels of several cytokines both
in vitro in primary microglia and
in vivo in TG mouse brain. This strongly supports the hypothesis that PPARδ treatment leads to reduced cytokine expression profile in multiple proinflammatory cytokines, which has beneficial effects on microglial activation and induces concomitant reduction in the brain Aβ burden. The fact that we detected similar changes both in primary microglia
in vitro and in 5XFAD mouse brain supports the contribution of microglia in the observed effects. A large number of cytokines and their corresponding receptors have been shown to be elevated in AD brain and the increased levels of many, such as IL-1β have detrimental effects on neuronal survival [
30]. CCL2 and CXCR2 are amongst inflammatory mediators the levels of which have been shown to be linked to neurodegeneration [
31-
34]. Similar to our study, different treatment paradigms in animal models of AD have demonstrated the link between reduction in the levels of CCL2 or CXCR2 and decreases in brain Aβ [
35-
38]. The fact that GW0742 induced a significant reduction in several measured cytokines implies that the potency of the drug relies on reducing the inflammatory milieu and inducing the concomitant beneficial actions in AD-like pathology.
In the current study the expression levels of C1qa and C3 were significantly elevated in the brains of 5XFAD mice and were reduced by GW0742 treatment. Most of the C1qa and C3 immunoreactivity was associated with GFAP immunoreactivity suggesting that although the level of astrocytic activation was not reduced, astrocytic expression of C1qa and C3 was diminished. The human AD brain has been shown to have increased expression of complement proteins [
39]. Aβ has been shown to directly activate the complement pathway [
40-
43] prompting the hypothesis that the activation of the complement pathway is detrimental in AD. Whereas microglial expression of complement proteins C1qa and C3 may enhance the Aβ phagocytic capacity [
44-
46], increased expression of these proteins has also been shown to lead to neuronal degeneration and death [
47-
49]. The overall decrease in C3 and C1qa in the current study may reflect a decrease in the brain inflammatory milieu. This may be beneficial for neuronal survival.
An important finding in our study is that treatment of AD mice with GW0742 not only reduced the brain inflammatory milieu but also prevented the loss in NeuN immunoreactivity in a specific brain area in 5XFAD mice that is susceptible to robust neuronal death [
25,
26]. We confirmed this finding
in vitro by using primary neuron-microglia co-cultures where GW0742 preserved MAP-2 immunoreactivity against inflammation-induced neuronal death. Although NeuN may not be a confirmative marker for neuronal death [
50] and our data do not represent absolute neuronal counts, our data are supported by studies showing the neuroprotective effect of PPARδ agonists in various models of neurodegenerative diseases [
13,
15-
17,
51] mainly via reducing inflammation and oxidative stress. It is noteworthy that GW0742 alone was not neuroprotective against glutamate exposure at the concentrations used in this study. The ability of GW0742 to provide direct neuroprotection
in vitro has yielded some contradictory results and may vary depending on treatment times, exposure concentrations and cellular models used and direct neuroprotection may require very high concentrations of PPARδ agonists [
13,
52,
53]. Our data imply that rather than providing direct neuroprotection against glutamate-induced neuronal death, GW0742 protects primary neurons from inflammation-induced neuronal death at the concentrations of GW0742 used in the study.
Our study shows for the first time that enhancement of PPARδ activity with a relatively short time window of only 2 weeks resulted in significant decrease in the amount of Aβ deposits and, importantly, it induced an overall decrease in the proinflammatory milieu slowing down the neuronal deterioration of 5XFAD mice. Our data warrant the activation of PPARδ as a potential therapeutic strategy to combat AD.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
TM and GL designed the study, TM, LD, MM and LN carried out the experiments, TM and GL wrote the manuscript. All authors read and approved the final manuscript.